Six weeks in and not a single undefeated NFL team remains. The Eagles and Bills both went down in Week 5, making this just the third time in two decades that every team has at least one loss before Week 6.

Advanced metrics show how little separates the league’s top teams. Seattle ranks first in DVOA yet sits just 4-2. The Rams and Commanders are top-five in efficiency but not in the standings. The Colts have backed up their numbers at 5-1, but even they’ve had close calls. Then there are the Buccaneers…5-1 with four wins that came when tied or trailing in the final minute.

Through five weeks, 24 games were won in the final two minutes or overtime, the most in league history through that span. Week 5 alone featured six double-digit comebacks.

As the season goes on, expect the teams built on true efficiency to separate from those surviving on variance. Strength of schedule will matter more and more. Fourth quarters will feel like a coin flip.

If you wanted parity in the NFL, this season is an example of what that looks like.

🏆 Think You Know Ball?

Prove It in the Props Battle.

$300 in prizes already claimed, and we’re just getting started.

Each week, the community locks in their picks. Get yours right and you’re in the running for $50 cash. Stack wins all season and the grand prize is a full year of Jaxon Premium ($999 value).

✅ Weekly $50 prizes
✅ Season-long leaderboard
✅ 100% free to enter

It’s bragging rights. It’s bankroll building. It’s all happening in NFL-Chat inside the Pine Sports Discord.

🏈 Thursday Night Football: Steelers vs. Bengals

The numbers tell a pretty clear story. Pittsburgh looks like a legitimate top-10 team by nearly every efficiency metric, while Cincinnati has cratered without Joe Burrow. The Steelers sit ninth in Total DVOA at +14.8%, powered by a defense ranked seventh overall and second in sacks. The Bengals, meanwhile, rank dead last in Total DVOA (–40.5%) and are giving up over 30 points per game.

The matchup tilts even further when you factor in pressure. Pittsburgh ranks fifth in QB pressure rate, and they’ll face a Bengals line that’s struggled all season to protect a backup quarterback. Cincinnati’s offense ranks 28th in efficiency, turning the ball over at one of the highest rates in football.

The market sees what the data does too. Pittsburgh opened as 5.5-point favorites, and our model gives them roughly a 70% chance to win outright. The data paints a consistent picture: this game should favor Pittsburgh in nearly every phase. Expect the Steelers to control the game through defense and field position. 

My Best Bet: Under 44.5 total points.

Pittsburgh’s elite defense and efficiency give them a clear edge over a struggling Bengals team missing Joe Burrow. So we asked Jaxon to generate the best player prop bets for this matchup, emphasizing Pittsburgh’s defensive dominance and Cincinnati’s offensive inefficiency under a backup QB.

DK Metcalf Over 58.5 Receiving Yards

Metcalf’s chemistry with Aaron Rodgers has quietly become one of the league’s most efficient duos. He’s averaging 71.2 yards per game, hitting this number in 60% of contests, and now faces a Bengals secondary ranked 27th in yards allowed to wideouts. With his 18.7 yards per catch average, he only needs a few solid connections to clear this total. Expect at least one deep shot to swing it.

Joe Flacco Over 0.5 Interceptions

Pittsburgh’s defense thrives on chaos and Flacco’s decision-making has fueled plenty of it. He’s thrown at least one pick in four of his last five games (1.2 per game average), while the Steelers rank 5th in total interceptions. Against a relentless pressure front that’s top-five in sacks, it’s hard to imagine Flacco escaping this one clean.

Joe Flacco Under 234.5 Passing Yards

Even with the Steelers’ secondary allowing yardage volume, Flacco hasn’t shown the efficiency to capitalize. He’s averaging just 206.8 passing yards per game and has gone under this mark in 80% of his starts. In his lone 2024 matchup with Pittsburgh, he threw for 168 yards. Unless this turns into unexpected garbage time, the under looks sharp.

Jaxon’s touchdown scorer picks are in. The red zone data points straight to two players the books clearly mispriced.

 🌲 The Pine Line

🏈 Over 80% of Survivor Pool entries are gone. We’re still standing. See which picks keep us alive in Week 7. 

⚖️ Six weeks in, and not one team is undefeated. It’s making Sundays a whole lot more interesting.

⚾️ I don’t know what’s crazier. Is it the play itself or that the call on the field was correct? 

🏆️ Two goals and another broken record in the books. Another ticket to the World Cup will have to wait. 

🏀 Enjoy watching the “Greek Freak” while you can. He isn’t sticking around any longer than he has to. 

Ready to Predict the NBA Season?

Building your own NBA model on Pine is simple and powerful. You’ll go through a few key steps: naming your model, choosing how much historical data to train it on, setting a rolling average window for current performance, and selecting the stats that drive your predictions.

Click here to head to the Pine NBA Model page. Once there, click “Build a New Model”, give your model a name, and choose what you want to predict for (Winner, Spread, or Total). For this example, we’ll be predicting a Winner, but the same steps apply to any option.

Choose Your Training Data Window

The first big decision is how far back to go when training your model. The goal is to balance data volume (more seasons = more samples) with data relevance (today’s NBA is faster and more three-point heavy than ever).

Recommended Range: 4 Seasons

Using data from the last four years gives your model enough volume to train effectively while keeping it representative of the modern game. Going back further than that risks introducing outdated trends that no longer apply to today’s style of play.

Set Your In-Season Performance Window

You’ll want to add a rolling average window, a dynamic measure of how teams are performing right now. This captures short-term form, recent player injuries, and lineup changes.

Recommended Range: 10–15 Games

10 Games reacts quickly to shifts in team dynamics, making it great for identifying hot streaks or slumps. 15 Games smooths out the randomness of small sample sizes for a more stable view of team strength. This balance between recency and stability helps your model stay current without overreacting to one-off results.

Select Your Predictive Stats

Now comes the fun part: choosing which stats power your predictions. Start with the core efficiency metrics that best represent how well a team scores, defends, and controls possessions. Here are some to consider:

Offensive Rating: Points scored per 100 possessions (offensive efficiency).

Defensive Rating: Points allowed per 100 possessions (defensive efficiency).

Effective Field Goal Percentage (eFG%): Adjusts for three-pointers being worth more than twos.

Turnover Rate (TOV%): How often a team turns the ball over per 100 possessions.

Rebounding Rates (OReb% & DReb%): Share of available rebounds secured.

Free Throw Rate (FTA/FGA): Frequency of getting to the line relative to shot attempts.

When you have what you want, click on build your model! You can view your model score and open up projections once the season starts back on the Model page.

Try different stats or settings for different results. Test, tweak, and improve just like the pros do. You can follow these same steps for NFL, NHL, and MLB, just look for the Predict tab on Pine and choose your sport!

Join the conversation!

Let us know if you’re tailing or who you think is due to find the endzone this weekend!

Tag us on X @PineSports_AI or join the Discord server so you don’t miss out!